We're sorry–what was the question?
Robert Enlow of the Friedman Foundation said this about polls some time ago:
One word makes a difference to public opinion and allows negative connotations to seep into documents meant to inform the public. That’s exactly what happened in the (annual Phi Delta Kappa / Gallup Poll of Public Attitudes Toward the Public Schools). As expected, it showed that the general public does not support school vouchers — 42 percent supported the idea.
Why was it expected?
For years, the word selection in the poll’s school-voucher question has been disputed. There have been rumblings that some of the connotation works to artificially lower support for school choice.
These rumblings prompted the Milton and Rose D. Friedman to carry out a study of its own to explore the potential of bias in the PDK poll. Conducted by leading research firm WirthlinWorldwide, the study, using a sound research methodology of a split-sample format, asked half of the participants the PDK question, while the other half was asked a more neutral school-voucher question.
Only a few words were changed.
For example, the PDK question asks, "do you favor or oppose allowing students and parents to choose a private school to attend at public expense?" The Friedman Foundation question read, "do you favor or oppose allowing students and parents to choose any school, public or private, to attend using public funds?"
Did the results raise questions of potential bias? We’ll let you draw your own conclusion.
The Friedman Foundation question netted support from 63 percent of Americans. The PDK question resulted in support from 41 percent. By changing only a few words, but keeping the meaning the same, support rose over 20 percent. (emphasis added)
If you think the PDK question is, well, questionable, then look at this St. Petersburg Times poll. The question:
"Some people say that state funds should only go to public schools. Others say the state of Florida should pay for private schooling if the public school a child attends is failing. In this situation, would you be for or against giving state funds to private schools?"
We don’t particularly have an issue with the first two sentences of that paragraph. But just who does the Times think it’s fooling with that third line? Not surprisingly, 61 percent said they were against.
Actually, this gives us an idea. So according to this, Gov. Jeb Bush is looking to spend 49 percent of next year’s budget on education. (Do not look any closer at that page, though–at least, not yet. We are, after all, trying to treat this like the Times treated that poll.) So! How about if the Alliance commissions the following poll question:
"Teachers’ unions have said Governor Jeb Bush needs to increase education spending. Since he has proposed spending nearly half of next year’s budget on education, do you believe that level of spending is good enough?"
Or how about this one?
"Governor Jeb Bush has proposed spending 49 percent of next year’s budget on education. Do you believe his spending proposal will benefit Florida schools?"
No, wait–we love this one:
"The state of Florida is planning to spend nearly half its annual budget on education. Do you believe that level of spending is too high?"
Now, look. We’re honest enough to realize that those poll questions are hopelessly slanted. Why? Let’s now go back to Bush’s budget numbers and actually, you know, tell the whole story by looking at the graph: that 49 percent number is of the general revenue fund, which doesn’t include lottery revenue, tobacco lawsuit money, and state trust or federal trust funds. Add those in, and education gets a total of 34 percent of all funding. Still the highest category, but it isn’t nearly half the total.
In short, no responsible pollster would use those sample questions and expect to get an accurate response. Which is why we find this poll by the Tampa Tribune a bit revealing in light of the Times poll.
Voters agree with the governor by a plurality, however, on another of his educational goals – reinstating the statewide private school tuition voucher program, which was struck down by the state Supreme Court.
The numbers (they’re buried in the next to last paragraph): 48 percent in favor/41 percent against. We can’t find the questions, but we’re willing to bet they’re a little more fair than those of the Times poll. Bottom line: you would be better off trusting an online poll than dreck like the St. Petersburg Times.
UPDATE: So saith Eduwonk.
The problem is the sentiment captured in the final line: "giving state funds to private schools." Sure, that’s part of the issue with vouchers but it’s not a clean way to ask the question because it loads it. I haven’t seen the entire poll so I don’t know if they asked the question in multiple ways but I’d bet that if you asked it with an emphasis on parents being able to chose schools, even private schools with public money, you’d get a different result.